Background and Significance Growth factor receptors are overexpressed in many cancers and their presence correlates with poor prognosis. Activation of growth factors signaling pathways results in increased proliferation, motility and resistance to chemo - and radiation therapies. It has been shown recently that the resistance of tumor cells to therapy aimed at blocking individual growth factor signaling pathways may be caused by cross-talk with another growth factor pathway. For example, activation of IGF pathway interferes with anti-EGF pathway therapy. Only very recently, appropriate software tools and sufficient computer power have become available that allow a quantitative computational exploration of signaling pathways as complex as the EGF/IGF pathway. We propose to use modeling and simulation software which has been developed specifically to allow quantitative exploration of complex signaling pathways. It is capable of automatically transforming sets of diagrammatical representations of bimolecular interactions within signaling pathways into quantitative simulations of the complete biochemical network. The software has successfully been used to predict previously unknown aspects of eukaryotic chemosensory G-protein coupled signaling pathway. After verification in the laboratory, the computer model applied to growth factor signaling pathways will provide means for a better understanding of their interactions and the combined effect on tumor response to radiation and chemotherapy facilitating design of an efficient multi-target approach to targeted therapy by blocking selected signaling molecules of these pathways. Experimental procedures Development of the model and quantitative simulations Based on available literature data, a detailed computational model of signaling pathways (initially EGF and IGF) will be created using SIMMUNE, a computer system for biochemical modeling and simulation of biochemical interactions. The model will be used to simulate cellular response to growth factors and computational exploration of possible therapeutic interventions aiming at selective blocking of these pathways. Other simulation techniques, including application of graph theory, boolenian networks, Mote Carlo simulations, etc, may be applied, depending on the problem at hand, availability of experimental data, and interest of our computational biology collaborators. In vitro verification of the model Prediction of the simulation will be compared with the effects of applied interventions on tumor cells such as gene expression associated with signaling pathways of interest and activation of selected proteins using molecular biology methods including western blots of signaling molecules, high density reverse-phase protein lysate microarrays, RT-PCR, and comparative gene expression analysis. Preference will be given to new quantitative methods. In vitro testing of the therapeutic efficacy of multitarget interference with EGF and IGF signaling pathways alone on in combination with radiation and chemotherapy The multitarget approach identified by the simulations as optimal for stand-alone or providing most potent sensitization of target breast tumor cells to radiation or chemotherapy will be tested in vitro using toxicity and clonogenic survival assays. These experiments will identify the optimal combinations and timing to be further tested in vivo. In vivo testing of the optimal combinations of therapeutic interventions Primary and metastatic breast cancer animal models will used to test the therapeutic efficacy of blocking signaling pathways for each receptor alone and in combination, as optimized by the computer simulations and validated by the in vitro testing. The effects of optimized molecular targeting alone or in combinations with chemo- and radiation therapy will be investigated. Project Status Simmune system is being modified to accommodate large complexes build during activation of EGFR. Database of relevant parameters have been build using the literature. Large discrepancies in some of the literature data necessitate their verification by our lab and collaborators in well controlled conditions. We are in the process of identification of the crucial parameters to be quantified using the state of the art methods. PathExplorer, a relatively simple approach, using graph theory to analyze signal transduction that allows to assess the effects of blocking (even partial, e.g. 25% or 50%) different nodes of signaling pathways on signal transduction and can identify combinations of nodes that have to be blocked in order to obtain an arbitrary combination of active and inactivated genes influenced by these pathways, is ready for verification. The findings have been presented as poster. Simulations of combined EGFR, HER2, IGF-1R, and IR are carried out. We are establishing a model for experimental verification of the simulations